Step-3.7-Flash

Name: Step-3.7-Flash
Author: stepfun-ai

by stepfun-ai

Apache-licensed multimodal MoE for image-text reasoning and fast visual QA

9.3Kdl/month

MoEparams

HuggingFace Run it on your own data, free

Identifiers

Model ID

stepfun-ai/Step-3.7-Flash

Feature URI

mixpeek://image_extractor@v1/stepfun_step37_flash_v1

Overview

Step 3.7 Flash is a new multimodal Mixture-of-Experts model from StepFun with image-text-to-text support. It is notable because the model card ships with Transformers and vLLM usage, making it more practical for teams that want a deployable open VLM rather than an API-only model.

On Mixpeek, Step 3.7 Flash is a candidate for scene captioning, visual question answering, screenshot analysis, and agent perception tasks where a single model needs to reason over images plus instructions.

Architecture

Vision-language Mixture-of-Experts model exposed through custom Transformers code and vLLM. Supports image-text chat prompts with Apache 2.0 licensing.

Mixpeek SDK Integration

// No extractor parameter takes a Hugging Face model id (checked against
// GET /v1/discovery/extractors, which returns 13), so Step-3.7-Flash runs
// on your side and the output is upserted through POST
// /v1/namespaces/{namespace_id}/documents/upsert. On Enterprise the other
// path is to upload the weights instead: POST /v1/namespaces/{id}/models
// accepts the huggingface format and a custom plugin loads them.
const res = await fetch(
  "https://api.mixpeek.com/v1/namespaces/ns_your_namespace/documents/upsert",
  {
    method: "POST",
    headers: {
      Authorization: "Bearer API_KEY",
      "Content-Type": "application/json",
    },
    body: JSON.stringify({
      collection_id: "col_your_collection",
      documents: [
        {
          document_id: "asset-00412",
          // The model produces text, so it lands in payload. Give the
          // collection a text vector index and embed that text to make it
          // searchable rather than only filterable.
          payload: { extracted_text: modelOutput, source_key: "archive/2026/asset-00412" },
          vectors: { "multimodal-embedding": embeddingOfModelOutput },
        },
      ],
    }),
  },
);

// Managed alternative, if this exact model is not the requirement:
// universal_extractor@v1 runs google/gemini-embedding-2
// (3072-d) over a bucket, with no inference of your own.